* Refactored reports to use pregenerated docs instead
* Fixed e2e test
* Fixed anothr bug
* Cleanup
* Fixed timezone conversion
* Updated ref files
* Updated reference files, refreshed ui/ux for report generation. Buttons toggle on and off based on if report exists
* Fixed linting errors, removed pytz
* Fixed test failure
* e2e test fix
* Speeding up unit tests
* Removed python time library that was causing performance issues with unit tests
* Updated poetry lock
* Unit test improvements
* Made change that ken reccomended
Note that this commit has some failing tests with it that also needed to be fixed; it is unclear why they are failing at the moment, though.
Signed-off-by: Carlo Costino <carlo.costino@gsa.gov>
- Deleted /stylesheets folder
- Removed sass build from gulpfile
- Changed gov links to usa links
- Changed other govuk styles, like breadcrumbs
- Changed name of uk_components file to us_components
- Fixed a few tests that broke on account of the changes
This changeset converts the display of dates and times to be just UTC to match the recent changes in the backend. This unwinds a bit of work that was done previously and allows us to start with a clean slate in how we want to approach displaying dates and times going forward. It also adds a bit of explanatory text to help users.
Signed-off-by: Carlo Costino <carlo.costino@gsa.gov>
Co-authored-by: stvnrlly <steven.reilly@gsa.gov>
I can’t find the changelog for this but it looks like somewhere someone
has decided that commas don’t need to be URL-encoded. This is true for
use in `href` attributes because it’s unambiguous that the comma is part
of the URL (unlike a closing quote for example, which could be
misinterpreted as HTML syntax).
This commit jut changes the test to reflect that the URLs generated by
Flask now have raw commas in them.
Some tests use the `client` fixture but don’t call any of its methods.
The reason for doing this is because the test depends on something in
the request context.
This commit replaces all those instances with `client_request`, which
also sets the request context.
These tests are the last ones that still use the `client` fixture. By
replacing it with `client_request` we will be able to say that no tests
should be using the `client` fixture directly.
We have a `client_request` fixture which does a bunch of useful stuff
like:
- checking the status code of the response
- returning a `BeautifulSoup` object
Lots of our tests still use an older fixture called `logged_in_client`.
This is not as good because:
- it returns a raw `Response` object
- doesn’t do the additional checks
- means our tests contain a lot of repetetive boilerplate like `page = BeautifulSoup(response.data.decode('utf-8'), 'html.parser')`
This commit converts all the tests using `logged_in_client` to:
use `client_request` instead.
Changes those fields in the following forms:
- SearchByNameForm
- SearchUsersByEmailForm
- SearchUsersForm
- SearchNotificationsForm
Includes changes to templates that use this form
and associated tests.
`EmailPreviewTemplate.subject` returns a string of HTML, with any
user-submitted HTML already escaped:
b5a61bfb7b/notifications_utils/template.py (L672)
What won’t be escaped is the HTML needed to redact the placeholders. We
generate this HTML so we know its safe, and doesn’t need to be escaped.
However when we pass it to Jinja, Jinja doesn’t know this, so will try
to escape it. This means users will see the raw HTML.
We can get around this by using Flask’s `Markup` class to tell Jinja
that the string is already sanitised and doesn’t need escaping again.
Text message templates don’t have this problem because they already
return `Markup`: b5a61bfb7b/notifications_utils/template.py (L288)
Letter templates don’t suffer from this problem (because they don’t
support redaction) but without making the same change they would still
double-escape ampersands, greater-than symbols, and so on.
This is what we do on the uploads page now. It makes it more obvious
why your search term has returned a certain result if you can see most
of the address, not just the first line.
We’re now normalising and storing the whole address in the
`normalised_to` field. Previously we were only storing the first line
of the address.
Enough time should now have passed that the field will have been
populated for all letters in the database.
Thus we can now tell users that it’s not just the first line they can
search by.
The search form makes a post request, so that phone numbers and email
addresses don’t show up in logs or browser history.
At most the API will return 50 results, with some pagination links. We
can’t easily give you links to click in the admin app, because links can
only perform get requests.
Because the value of seeing more than 50 results feels quite low (users
will probably make their search more specific before scrolling through
all 50) let’s just show a message saying only the first 50 results are
displayed.
Like we have search by email address or phone number, finding an
individual letter is a common task. At the moment users are having to
click through pages and pages of letters to find the one they’re looking
for.
Users of the API will also be able to search by reference, same as for
emails and text messages. But we only show this hint text to users who
have some API keys.
